Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7f7e1f933668>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7f7e1f816710>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.1.0
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    # TODO: Implement Function
    input_real = tf.placeholder(tf.float32, 
                                shape=(None, image_width, image_height, image_channels),
                                name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')
    learning_rate = tf.placeholder(tf.float32, name='learning_rate')

    return input_real, inputs_z, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

alpha = 0.2
def leaky_relu(x):
    return tf.maximum(alpha*x,x)

def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    # TODO: Implement Function
    with tf.variable_scope('discriminator', reuse=reuse):
        conv_layer1 = tf.layers.conv2d(images, 64, 5, strides=2, padding='same')
        relu1 = leaky_relu(conv_layer1)
        
        conv_layer2 = tf.layers.conv2d(conv_layer1, 128, 5, strides=2, padding='same')
        conv_layer2 = tf.layers.batch_normalization(conv_layer2, training=True)
        relu2 = leaky_relu(conv_layer2)
        
        conv_layer3 = tf.layers.conv2d(conv_layer2, 256, 5, strides=2, padding='same')
        conv_layer3 = tf.layers.batch_normalization(conv_layer3, training=True)
        relu3 = leaky_relu(conv_layer3)
        
        conv_layer4 = tf.layers.conv2d(conv_layer3, 512, 5, strides=2, padding='same')
        conv_layer4 = tf.layers.batch_normalization(conv_layer4)
        relu4 = leaky_relu(conv_layer4)
        
        #Flat
        flat = tf.reshape(relu3, (-1, 4*4*256))
        
        #Logits
        logits = tf.layers.dense(flat, 1)
        out = tf.sigmoid(logits)

    return out, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    # TODO: Implement Function
    with tf.variable_scope('generator', reuse=False if is_train==True else True):
        #Fully connected layer
        fully_conn = tf.layers.dense(z, 7*7*512)
        fully_conn = tf.reshape(fully_conn, (-1,7,7,512))
        fully_conn = tf.layers.batch_normalization(fully_conn, training=is_train)
        fully_conn = leaky_relu(fully_conn)
        # 7x7x512

        #Convolutions
        conv_layer1 = tf.layers.conv2d_transpose(fully_conn, 256, 3, strides=2, padding='same')
        conv_layer1 = tf.layers.batch_normalization(conv_layer1, training=is_train)
        conv_layer1 = leaky_relu(conv_layer1)
        # 14x14x256

        conv_layer2 = tf.layers.conv2d_transpose(conv_layer1, 128, 3, strides=1, padding='same')
        conv_layer2 = tf.layers.batch_normalization(conv_layer2, training=is_train)
        conv_layer2 = leaky_relu(conv_layer2)
        # 14x14x128

        conv_layer3 = tf.layers.conv2d_transpose(conv_layer2, 64, 3, strides=2, padding='same')
        conv_layer3 = tf.layers.batch_normalization(conv_layer3, training=is_train)
        conv_layer3 = leaky_relu(conv_layer3)
        # 28x28x64   

        # Output layer
        logits = tf.layers.conv2d_transpose(conv_layer3, out_channel_dim, 3, strides=1, padding='same')
        # 28x28xout_dim 

        gen_out = tf.tanh(logits)

    return gen_out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    # TODO: Implement Function
    # Label smoothing 
    smooth = 0.1
    
    g_model = generator(input_z, out_channel_dim, is_train=True)
    d_model_real, d_logits_real = discriminator(input_real, reuse=False)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)
    
    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real,
                                                labels=(tf.ones_like(d_model_real)*(1-smooth)) ))
    
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake,
                                                labels=tf.zeros_like(d_model_fake)))

    dicriminator_loss = d_loss_real + d_loss_fake
    
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    
    return dicriminator_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # TODO: Implement Function
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)
    
    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # TODO: Build Model
    print("Shape:", data_shape)
    print("Learning rate:", learning_rate)
    img_width = data_shape[1]
    img_height = data_shape[2]
    img_channels = data_shape[3]
    
    # Init GAN
    input_real, input_z, lr = model_inputs(img_width, img_height, img_channels, z_dim)
    print(input_real)
    print(input_z)
    d_loss, g_loss = model_loss(input_real, input_z, img_channels)
    d_opt, g_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
    
    # Train
    steps = 0
    saver = tf.train.Saver()    
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                # TODO: Train Model
                steps += 1
                
                #adjust the range
                batch_images *= 2
                
                batch_z = np.random.uniform(-1,1, size=(batch_size, z_dim))
                
                _ = sess.run(d_opt, feed_dict={input_real: batch_images, input_z: batch_z, lr: learning_rate})
                _ = sess.run(g_opt, feed_dict={input_real: batch_images, input_z: batch_z, lr: learning_rate})
                
                if steps % 5 == 0:
                    d_loss_train = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    g_loss_train = g_loss.eval({input_z: batch_z})
                    
                    print("Epoch {}/{}...".format(epoch_i+1, epoch_count),
                         "Discriminator Loss: {:.4f}...".format(d_loss_train),
                         "Generator Loss: {:.4f}".format(g_loss_train))
                    
                if steps % 100 == 0:
                    show_generator_output(sess, 25, input_z, img_channels, data_image_mode)

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 64
z_dim = 100
learning_rate = 0.0002
beta1 = 0.5
alpha = 0.2

tf.reset_default_graph()
"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 3

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))

with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

Shape: (60000, 28, 28, 1)
Learning rate: 0.0002
Tensor("input_real:0", shape=(?, 28, 28, 1), dtype=float32)
Tensor("input_z:0", shape=(?, 100), dtype=float32)
Epoch 1/3... Discriminator Loss: 6.0822... Generator Loss: 0.0037
Epoch 1/3... Discriminator Loss: 6.0905... Generator Loss: 0.0052
Epoch 1/3... Discriminator Loss: 5.1466... Generator Loss: 0.0116
Epoch 1/3... Discriminator Loss: 0.5219... Generator Loss: 2.2454
Epoch 1/3... Discriminator Loss: 4.3805... Generator Loss: 0.0237
Epoch 1/3... Discriminator Loss: 0.3559... Generator Loss: 6.9227
Epoch 1/3... Discriminator Loss: 0.3841... Generator Loss: 3.4970
Epoch 1/3... Discriminator Loss: 0.3621... Generator Loss: 4.0656
Epoch 1/3... Discriminator Loss: 0.3437... Generator Loss: 11.4582
Epoch 1/3... Discriminator Loss: 0.4163... Generator Loss: 2.7562
Epoch 1/3... Discriminator Loss: 0.3992... Generator Loss: 2.8926
Epoch 1/3... Discriminator Loss: 0.5545... Generator Loss: 5.6646
Epoch 1/3... Discriminator Loss: 0.4015... Generator Loss: 3.3633
Epoch 1/3... Discriminator Loss: 1.4801... Generator Loss: 0.4955
Epoch 1/3... Discriminator Loss: 0.5072... Generator Loss: 4.3923
Epoch 1/3... Discriminator Loss: 0.3537... Generator Loss: 6.2240
Epoch 1/3... Discriminator Loss: 1.8052... Generator Loss: 0.5104
Epoch 1/3... Discriminator Loss: 0.7781... Generator Loss: 1.2836
Epoch 1/3... Discriminator Loss: 0.8335... Generator Loss: 1.2480
Epoch 1/3... Discriminator Loss: 0.5785... Generator Loss: 2.1525

Epoch 1/3... Discriminator Loss: 0.9029... Generator Loss: 1.0882
Epoch 1/3... Discriminator Loss: 0.9764... Generator Loss: 1.2225
Epoch 1/3... Discriminator Loss: 0.8071... Generator Loss: 1.2214
Epoch 1/3... Discriminator Loss: 0.6930... Generator Loss: 1.4470
Epoch 1/3... Discriminator Loss: 1.1244... Generator Loss: 0.9557
Epoch 1/3... Discriminator Loss: 0.8019... Generator Loss: 1.5602
Epoch 1/3... Discriminator Loss: 1.9883... Generator Loss: 0.3920
Epoch 1/3... Discriminator Loss: 1.0820... Generator Loss: 1.1834
Epoch 1/3... Discriminator Loss: 1.7256... Generator Loss: 0.3621
Epoch 1/3... Discriminator Loss: 1.9771... Generator Loss: 0.5938
Epoch 1/3... Discriminator Loss: 1.2074... Generator Loss: 0.8041
Epoch 1/3... Discriminator Loss: 1.4878... Generator Loss: 1.7054
Epoch 1/3... Discriminator Loss: 1.2966... Generator Loss: 0.8153
Epoch 1/3... Discriminator Loss: 1.3762... Generator Loss: 0.6862
Epoch 1/3... Discriminator Loss: 1.0592... Generator Loss: 1.0182
Epoch 1/3... Discriminator Loss: 1.5238... Generator Loss: 0.7505
Epoch 1/3... Discriminator Loss: 1.0842... Generator Loss: 1.2125
Epoch 1/3... Discriminator Loss: 1.5017... Generator Loss: 1.6108
Epoch 1/3... Discriminator Loss: 1.3837... Generator Loss: 1.0999
Epoch 1/3... Discriminator Loss: 1.2163... Generator Loss: 0.9069

Epoch 1/3... Discriminator Loss: 1.3221... Generator Loss: 0.9221
Epoch 1/3... Discriminator Loss: 1.4342... Generator Loss: 1.1022
Epoch 1/3... Discriminator Loss: 1.5856... Generator Loss: 0.5897
Epoch 1/3... Discriminator Loss: 1.4770... Generator Loss: 0.6105
Epoch 1/3... Discriminator Loss: 1.5221... Generator Loss: 0.5760
Epoch 1/3... Discriminator Loss: 1.3557... Generator Loss: 0.7356
Epoch 1/3... Discriminator Loss: 1.3648... Generator Loss: 0.8210
Epoch 1/3... Discriminator Loss: 1.2522... Generator Loss: 0.9171
Epoch 1/3... Discriminator Loss: 1.2526... Generator Loss: 0.8441
Epoch 1/3... Discriminator Loss: 1.3756... Generator Loss: 1.0233
Epoch 1/3... Discriminator Loss: 1.3939... Generator Loss: 0.9968
Epoch 1/3... Discriminator Loss: 1.4352... Generator Loss: 0.5685
Epoch 1/3... Discriminator Loss: 1.4148... Generator Loss: 0.8142
Epoch 1/3... Discriminator Loss: 1.3421... Generator Loss: 0.7412
Epoch 1/3... Discriminator Loss: 1.3516... Generator Loss: 1.0129
Epoch 1/3... Discriminator Loss: 1.2855... Generator Loss: 0.7629
Epoch 1/3... Discriminator Loss: 1.4498... Generator Loss: 0.6590
Epoch 1/3... Discriminator Loss: 1.2493... Generator Loss: 1.0339
Epoch 1/3... Discriminator Loss: 1.3799... Generator Loss: 0.8189
Epoch 1/3... Discriminator Loss: 1.3674... Generator Loss: 0.6274

Epoch 1/3... Discriminator Loss: 1.3991... Generator Loss: 0.6765
Epoch 1/3... Discriminator Loss: 1.2828... Generator Loss: 0.7801
Epoch 1/3... Discriminator Loss: 1.3470... Generator Loss: 0.7214
Epoch 1/3... Discriminator Loss: 1.3541... Generator Loss: 0.7674
Epoch 1/3... Discriminator Loss: 1.2457... Generator Loss: 0.8534
Epoch 1/3... Discriminator Loss: 1.4048... Generator Loss: 0.7517
Epoch 1/3... Discriminator Loss: 1.2894... Generator Loss: 0.7389
Epoch 1/3... Discriminator Loss: 1.3642... Generator Loss: 0.6967
Epoch 1/3... Discriminator Loss: 1.3444... Generator Loss: 0.9521
Epoch 1/3... Discriminator Loss: 1.3199... Generator Loss: 1.0001
Epoch 1/3... Discriminator Loss: 1.3911... Generator Loss: 0.7341
Epoch 1/3... Discriminator Loss: 1.3649... Generator Loss: 0.6568
Epoch 1/3... Discriminator Loss: 1.3327... Generator Loss: 0.8253
Epoch 1/3... Discriminator Loss: 1.3335... Generator Loss: 0.7805
Epoch 1/3... Discriminator Loss: 1.2180... Generator Loss: 0.9954
Epoch 1/3... Discriminator Loss: 1.3825... Generator Loss: 0.8383
Epoch 1/3... Discriminator Loss: 1.3432... Generator Loss: 0.6755
Epoch 1/3... Discriminator Loss: 1.2837... Generator Loss: 0.8964
Epoch 1/3... Discriminator Loss: 1.2621... Generator Loss: 1.0894
Epoch 1/3... Discriminator Loss: 1.2883... Generator Loss: 0.9690

Epoch 1/3... Discriminator Loss: 1.2802... Generator Loss: 0.7778
Epoch 1/3... Discriminator Loss: 1.3586... Generator Loss: 0.7886
Epoch 1/3... Discriminator Loss: 1.2202... Generator Loss: 0.7653
Epoch 1/3... Discriminator Loss: 1.1832... Generator Loss: 0.9075
Epoch 1/3... Discriminator Loss: 1.2226... Generator Loss: 0.8506
Epoch 1/3... Discriminator Loss: 1.2669... Generator Loss: 0.9455
Epoch 1/3... Discriminator Loss: 1.2384... Generator Loss: 1.0066
Epoch 1/3... Discriminator Loss: 1.2742... Generator Loss: 0.7807
Epoch 1/3... Discriminator Loss: 1.2439... Generator Loss: 0.9864
Epoch 1/3... Discriminator Loss: 1.2070... Generator Loss: 0.9361
Epoch 1/3... Discriminator Loss: 1.2198... Generator Loss: 0.8669
Epoch 1/3... Discriminator Loss: 1.2914... Generator Loss: 0.7406
Epoch 1/3... Discriminator Loss: 1.2565... Generator Loss: 0.9403
Epoch 1/3... Discriminator Loss: 1.2308... Generator Loss: 0.8802
Epoch 1/3... Discriminator Loss: 1.1822... Generator Loss: 0.9247
Epoch 1/3... Discriminator Loss: 1.1953... Generator Loss: 0.8113
Epoch 1/3... Discriminator Loss: 1.2065... Generator Loss: 0.9724
Epoch 1/3... Discriminator Loss: 1.1724... Generator Loss: 0.9989
Epoch 1/3... Discriminator Loss: 1.1676... Generator Loss: 1.0549
Epoch 1/3... Discriminator Loss: 1.2644... Generator Loss: 0.7621

Epoch 1/3... Discriminator Loss: 1.2704... Generator Loss: 0.8454
Epoch 1/3... Discriminator Loss: 1.1862... Generator Loss: 0.9737
Epoch 1/3... Discriminator Loss: 1.1935... Generator Loss: 0.9377
Epoch 1/3... Discriminator Loss: 1.1496... Generator Loss: 1.0704
Epoch 1/3... Discriminator Loss: 1.2642... Generator Loss: 1.1472
Epoch 1/3... Discriminator Loss: 1.1790... Generator Loss: 0.8943
Epoch 1/3... Discriminator Loss: 1.2441... Generator Loss: 0.9227
Epoch 1/3... Discriminator Loss: 1.1837... Generator Loss: 0.9019
Epoch 1/3... Discriminator Loss: 1.1124... Generator Loss: 1.0181
Epoch 1/3... Discriminator Loss: 1.1796... Generator Loss: 1.1385
Epoch 1/3... Discriminator Loss: 1.1935... Generator Loss: 1.0121
Epoch 1/3... Discriminator Loss: 1.1597... Generator Loss: 1.2439
Epoch 1/3... Discriminator Loss: 1.1815... Generator Loss: 1.0395
Epoch 1/3... Discriminator Loss: 1.2423... Generator Loss: 1.0711
Epoch 1/3... Discriminator Loss: 1.2568... Generator Loss: 0.8193
Epoch 1/3... Discriminator Loss: 1.1942... Generator Loss: 0.9573
Epoch 1/3... Discriminator Loss: 1.1760... Generator Loss: 0.8863
Epoch 1/3... Discriminator Loss: 1.1000... Generator Loss: 1.0335
Epoch 1/3... Discriminator Loss: 1.2177... Generator Loss: 0.9247
Epoch 1/3... Discriminator Loss: 1.1380... Generator Loss: 0.9031

Epoch 1/3... Discriminator Loss: 1.2888... Generator Loss: 0.8223
Epoch 1/3... Discriminator Loss: 1.2285... Generator Loss: 0.8973
Epoch 1/3... Discriminator Loss: 1.2082... Generator Loss: 1.0183
Epoch 1/3... Discriminator Loss: 1.1684... Generator Loss: 1.0480
Epoch 1/3... Discriminator Loss: 1.1943... Generator Loss: 1.0607
Epoch 1/3... Discriminator Loss: 1.2665... Generator Loss: 0.7175
Epoch 1/3... Discriminator Loss: 1.1552... Generator Loss: 1.2632
Epoch 1/3... Discriminator Loss: 1.1618... Generator Loss: 1.2210
Epoch 1/3... Discriminator Loss: 1.1714... Generator Loss: 0.7840
Epoch 1/3... Discriminator Loss: 1.0893... Generator Loss: 1.0710
Epoch 1/3... Discriminator Loss: 1.2411... Generator Loss: 0.7691
Epoch 1/3... Discriminator Loss: 1.2070... Generator Loss: 0.8297
Epoch 1/3... Discriminator Loss: 1.0703... Generator Loss: 1.0271
Epoch 1/3... Discriminator Loss: 1.1359... Generator Loss: 1.0561
Epoch 1/3... Discriminator Loss: 1.0986... Generator Loss: 0.9710
Epoch 1/3... Discriminator Loss: 1.1564... Generator Loss: 0.9606
Epoch 1/3... Discriminator Loss: 1.1720... Generator Loss: 1.2299
Epoch 1/3... Discriminator Loss: 1.1524... Generator Loss: 1.0919
Epoch 1/3... Discriminator Loss: 1.0923... Generator Loss: 0.9804
Epoch 1/3... Discriminator Loss: 1.1552... Generator Loss: 0.8332

Epoch 1/3... Discriminator Loss: 1.1379... Generator Loss: 1.1651
Epoch 1/3... Discriminator Loss: 1.1446... Generator Loss: 1.0296
Epoch 1/3... Discriminator Loss: 1.0936... Generator Loss: 1.0654
Epoch 1/3... Discriminator Loss: 1.0932... Generator Loss: 1.0345
Epoch 1/3... Discriminator Loss: 1.1241... Generator Loss: 0.9524
Epoch 1/3... Discriminator Loss: 1.1833... Generator Loss: 0.9402
Epoch 1/3... Discriminator Loss: 1.2097... Generator Loss: 0.7752
Epoch 1/3... Discriminator Loss: 1.1157... Generator Loss: 0.8773
Epoch 1/3... Discriminator Loss: 1.1207... Generator Loss: 0.8131
Epoch 1/3... Discriminator Loss: 1.1101... Generator Loss: 0.9801
Epoch 1/3... Discriminator Loss: 1.1219... Generator Loss: 1.0599
Epoch 1/3... Discriminator Loss: 1.1626... Generator Loss: 0.9398
Epoch 1/3... Discriminator Loss: 1.1466... Generator Loss: 0.9743
Epoch 1/3... Discriminator Loss: 1.0873... Generator Loss: 1.0488
Epoch 1/3... Discriminator Loss: 1.1084... Generator Loss: 0.9987
Epoch 1/3... Discriminator Loss: 1.0832... Generator Loss: 1.2452
Epoch 1/3... Discriminator Loss: 1.2150... Generator Loss: 0.9172
Epoch 1/3... Discriminator Loss: 1.1563... Generator Loss: 1.0183
Epoch 1/3... Discriminator Loss: 1.1295... Generator Loss: 0.9767
Epoch 1/3... Discriminator Loss: 1.1890... Generator Loss: 1.1316

Epoch 1/3... Discriminator Loss: 1.1877... Generator Loss: 0.9734
Epoch 1/3... Discriminator Loss: 1.1643... Generator Loss: 0.8495
Epoch 1/3... Discriminator Loss: 1.1053... Generator Loss: 1.0885
Epoch 1/3... Discriminator Loss: 1.1777... Generator Loss: 0.7547
Epoch 1/3... Discriminator Loss: 1.1543... Generator Loss: 0.8011
Epoch 1/3... Discriminator Loss: 1.1937... Generator Loss: 1.0675
Epoch 1/3... Discriminator Loss: 1.2200... Generator Loss: 0.8180
Epoch 1/3... Discriminator Loss: 1.1860... Generator Loss: 1.0418
Epoch 1/3... Discriminator Loss: 1.1639... Generator Loss: 1.1522
Epoch 1/3... Discriminator Loss: 1.1656... Generator Loss: 0.9395
Epoch 1/3... Discriminator Loss: 1.1755... Generator Loss: 1.1045
Epoch 1/3... Discriminator Loss: 1.0976... Generator Loss: 1.0851
Epoch 1/3... Discriminator Loss: 1.1752... Generator Loss: 0.9984
Epoch 1/3... Discriminator Loss: 1.1130... Generator Loss: 0.9913
Epoch 1/3... Discriminator Loss: 1.1541... Generator Loss: 0.8342
Epoch 1/3... Discriminator Loss: 1.1044... Generator Loss: 0.9413
Epoch 1/3... Discriminator Loss: 1.0828... Generator Loss: 1.1198
Epoch 1/3... Discriminator Loss: 1.1000... Generator Loss: 1.0256
Epoch 1/3... Discriminator Loss: 1.1582... Generator Loss: 0.9397
Epoch 1/3... Discriminator Loss: 1.2709... Generator Loss: 0.7134

Epoch 1/3... Discriminator Loss: 1.1046... Generator Loss: 0.9978
Epoch 1/3... Discriminator Loss: 1.1003... Generator Loss: 1.1755
Epoch 1/3... Discriminator Loss: 1.3237... Generator Loss: 0.6796
Epoch 1/3... Discriminator Loss: 1.1230... Generator Loss: 1.0540
Epoch 1/3... Discriminator Loss: 1.1725... Generator Loss: 0.9492
Epoch 1/3... Discriminator Loss: 1.1507... Generator Loss: 1.0714
Epoch 1/3... Discriminator Loss: 1.1849... Generator Loss: 0.9876
Epoch 2/3... Discriminator Loss: 1.1433... Generator Loss: 0.8538
Epoch 2/3... Discriminator Loss: 1.2806... Generator Loss: 0.8714
Epoch 2/3... Discriminator Loss: 1.1948... Generator Loss: 0.9774
Epoch 2/3... Discriminator Loss: 1.2077... Generator Loss: 0.8392
Epoch 2/3... Discriminator Loss: 1.2475... Generator Loss: 1.0571
Epoch 2/3... Discriminator Loss: 1.2476... Generator Loss: 0.9308
Epoch 2/3... Discriminator Loss: 1.1719... Generator Loss: 0.8803
Epoch 2/3... Discriminator Loss: 1.2001... Generator Loss: 0.8998
Epoch 2/3... Discriminator Loss: 1.2071... Generator Loss: 1.0400
Epoch 2/3... Discriminator Loss: 1.2845... Generator Loss: 0.6687
Epoch 2/3... Discriminator Loss: 1.2756... Generator Loss: 0.8020
Epoch 2/3... Discriminator Loss: 1.2493... Generator Loss: 1.1815
Epoch 2/3... Discriminator Loss: 1.1799... Generator Loss: 1.0429

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 64
z_dim = 100
learning_rate = 0.0002
beta1 = 0.2
alpha = 0.08

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 8

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Shape: (202599, 28, 28, 3)
Learning rate: 0.0002
Tensor("input_real:0", shape=(?, 28, 28, 3), dtype=float32)
Tensor("input_z:0", shape=(?, 100), dtype=float32)
Epoch 1/8... Discriminator Loss: 2.4932... Generator Loss: 0.1610
Epoch 1/8... Discriminator Loss: 2.1317... Generator Loss: 0.2110
Epoch 1/8... Discriminator Loss: 3.9330... Generator Loss: 0.0359
Epoch 1/8... Discriminator Loss: 1.3455... Generator Loss: 0.6105
Epoch 1/8... Discriminator Loss: 0.9140... Generator Loss: 1.0055
Epoch 1/8... Discriminator Loss: 0.8036... Generator Loss: 1.1480
Epoch 1/8... Discriminator Loss: 0.7104... Generator Loss: 1.3240
Epoch 1/8... Discriminator Loss: 2.5623... Generator Loss: 0.1350
Epoch 1/8... Discriminator Loss: 0.9775... Generator Loss: 1.3277
Epoch 1/8... Discriminator Loss: 3.2330... Generator Loss: 0.0982
Epoch 1/8... Discriminator Loss: 1.7917... Generator Loss: 0.4285
Epoch 1/8... Discriminator Loss: 1.5159... Generator Loss: 0.7066
Epoch 1/8... Discriminator Loss: 1.5382... Generator Loss: 0.4320
Epoch 1/8... Discriminator Loss: 1.4265... Generator Loss: 1.7554
Epoch 1/8... Discriminator Loss: 1.5612... Generator Loss: 0.4957
Epoch 1/8... Discriminator Loss: 1.2405... Generator Loss: 1.2701
Epoch 1/8... Discriminator Loss: 1.9767... Generator Loss: 0.2910
Epoch 1/8... Discriminator Loss: 1.8935... Generator Loss: 2.0352
Epoch 1/8... Discriminator Loss: 2.0190... Generator Loss: 0.2271
Epoch 1/8... Discriminator Loss: 1.5136... Generator Loss: 1.4527

Epoch 1/8... Discriminator Loss: 1.5815... Generator Loss: 0.4022
Epoch 1/8... Discriminator Loss: 1.6513... Generator Loss: 1.5575
Epoch 1/8... Discriminator Loss: 1.4737... Generator Loss: 0.4782
Epoch 1/8... Discriminator Loss: 1.3762... Generator Loss: 1.5433
Epoch 1/8... Discriminator Loss: 1.7064... Generator Loss: 0.3573
Epoch 1/8... Discriminator Loss: 1.4286... Generator Loss: 1.2672
Epoch 1/8... Discriminator Loss: 1.6830... Generator Loss: 0.3422
Epoch 1/8... Discriminator Loss: 1.2554... Generator Loss: 1.6057
Epoch 1/8... Discriminator Loss: 0.9129... Generator Loss: 1.0918
Epoch 1/8... Discriminator Loss: 1.4548... Generator Loss: 2.5926
Epoch 1/8... Discriminator Loss: 1.4382... Generator Loss: 0.4672
Epoch 1/8... Discriminator Loss: 1.0843... Generator Loss: 3.0134
Epoch 1/8... Discriminator Loss: 1.7088... Generator Loss: 0.3598
Epoch 1/8... Discriminator Loss: 0.8436... Generator Loss: 1.5302
Epoch 1/8... Discriminator Loss: 0.7433... Generator Loss: 1.3263
Epoch 1/8... Discriminator Loss: 1.2778... Generator Loss: 3.5281
Epoch 1/8... Discriminator Loss: 1.2086... Generator Loss: 0.6560
Epoch 1/8... Discriminator Loss: 1.0050... Generator Loss: 0.8204
Epoch 1/8... Discriminator Loss: 0.5001... Generator Loss: 2.3793
Epoch 1/8... Discriminator Loss: 0.7416... Generator Loss: 1.2659

Epoch 1/8... Discriminator Loss: 0.6045... Generator Loss: 2.7066
Epoch 1/8... Discriminator Loss: 1.0983... Generator Loss: 0.7021
Epoch 1/8... Discriminator Loss: 0.5353... Generator Loss: 2.6021
Epoch 1/8... Discriminator Loss: 1.3699... Generator Loss: 0.4829
Epoch 1/8... Discriminator Loss: 1.4529... Generator Loss: 1.8569
Epoch 1/8... Discriminator Loss: 1.0691... Generator Loss: 3.1131
Epoch 1/8... Discriminator Loss: 0.6125... Generator Loss: 1.8144
Epoch 1/8... Discriminator Loss: 0.6227... Generator Loss: 3.6537
Epoch 1/8... Discriminator Loss: 1.3095... Generator Loss: 0.5313
Epoch 1/8... Discriminator Loss: 0.6893... Generator Loss: 1.3883
Epoch 1/8... Discriminator Loss: 0.6138... Generator Loss: 3.3494
Epoch 1/8... Discriminator Loss: 0.5478... Generator Loss: 1.9602
Epoch 1/8... Discriminator Loss: 0.4511... Generator Loss: 2.9204
Epoch 1/8... Discriminator Loss: 0.4382... Generator Loss: 3.0086
Epoch 1/8... Discriminator Loss: 0.4965... Generator Loss: 3.0497
Epoch 1/8... Discriminator Loss: 0.6833... Generator Loss: 1.3522
Epoch 1/8... Discriminator Loss: 1.2420... Generator Loss: 3.7890
Epoch 1/8... Discriminator Loss: 0.4646... Generator Loss: 2.6278
Epoch 1/8... Discriminator Loss: 0.4777... Generator Loss: 2.3260
Epoch 1/8... Discriminator Loss: 0.4654... Generator Loss: 4.4528

Epoch 1/8... Discriminator Loss: 0.5616... Generator Loss: 1.8744
Epoch 1/8... Discriminator Loss: 0.3808... Generator Loss: 4.2102
Epoch 1/8... Discriminator Loss: 0.5538... Generator Loss: 1.8796
Epoch 1/8... Discriminator Loss: 0.4850... Generator Loss: 4.0952
Epoch 1/8... Discriminator Loss: 0.3925... Generator Loss: 3.3483
Epoch 1/8... Discriminator Loss: 0.5358... Generator Loss: 4.8815
Epoch 1/8... Discriminator Loss: 0.3673... Generator Loss: 4.0647
Epoch 1/8... Discriminator Loss: 0.4811... Generator Loss: 5.3786
Epoch 1/8... Discriminator Loss: 0.7648... Generator Loss: 1.2035
Epoch 1/8... Discriminator Loss: 0.7138... Generator Loss: 5.0592
Epoch 1/8... Discriminator Loss: 1.8548... Generator Loss: 5.1378
Epoch 1/8... Discriminator Loss: 1.6398... Generator Loss: 0.3568
Epoch 1/8... Discriminator Loss: 0.7530... Generator Loss: 1.8801
Epoch 1/8... Discriminator Loss: 0.7840... Generator Loss: 1.2386
Epoch 1/8... Discriminator Loss: 0.9964... Generator Loss: 4.6712
Epoch 1/8... Discriminator Loss: 0.4281... Generator Loss: 5.4229
Epoch 1/8... Discriminator Loss: 0.4712... Generator Loss: 6.5111
Epoch 1/8... Discriminator Loss: 0.4452... Generator Loss: 2.8121
Epoch 1/8... Discriminator Loss: 0.8576... Generator Loss: 3.4411
Epoch 1/8... Discriminator Loss: 0.4599... Generator Loss: 5.2417

Epoch 1/8... Discriminator Loss: 0.3900... Generator Loss: 3.5271
Epoch 1/8... Discriminator Loss: 0.5987... Generator Loss: 1.6915
Epoch 1/8... Discriminator Loss: 1.4887... Generator Loss: 3.8948
Epoch 1/8... Discriminator Loss: 1.0371... Generator Loss: 0.7756
Epoch 1/8... Discriminator Loss: 0.4394... Generator Loss: 3.9671
Epoch 1/8... Discriminator Loss: 0.4816... Generator Loss: 3.8934
Epoch 1/8... Discriminator Loss: 0.4120... Generator Loss: 2.9916
Epoch 1/8... Discriminator Loss: 0.3607... Generator Loss: 5.2824
Epoch 1/8... Discriminator Loss: 0.3526... Generator Loss: 5.8618
Epoch 1/8... Discriminator Loss: 0.4372... Generator Loss: 2.6630
Epoch 1/8... Discriminator Loss: 0.3944... Generator Loss: 3.1737
Epoch 1/8... Discriminator Loss: 1.7752... Generator Loss: 4.7955
Epoch 1/8... Discriminator Loss: 0.7070... Generator Loss: 1.3757
Epoch 1/8... Discriminator Loss: 0.5647... Generator Loss: 3.8721
Epoch 1/8... Discriminator Loss: 0.6774... Generator Loss: 1.5083
Epoch 1/8... Discriminator Loss: 0.7188... Generator Loss: 5.6647
Epoch 1/8... Discriminator Loss: 0.4361... Generator Loss: 3.0787
Epoch 1/8... Discriminator Loss: 0.9687... Generator Loss: 5.9411
Epoch 1/8... Discriminator Loss: 0.3679... Generator Loss: 4.2945
Epoch 1/8... Discriminator Loss: 0.4566... Generator Loss: 4.3698

Epoch 1/8... Discriminator Loss: 0.3968... Generator Loss: 3.5595
Epoch 1/8... Discriminator Loss: 1.4217... Generator Loss: 3.6908
Epoch 1/8... Discriminator Loss: 0.3736... Generator Loss: 3.7876
Epoch 1/8... Discriminator Loss: 0.4396... Generator Loss: 3.2361
Epoch 1/8... Discriminator Loss: 2.0522... Generator Loss: 4.2414
Epoch 1/8... Discriminator Loss: 0.6817... Generator Loss: 1.4578
Epoch 1/8... Discriminator Loss: 0.5396... Generator Loss: 3.6888
Epoch 1/8... Discriminator Loss: 0.4655... Generator Loss: 2.6487
Epoch 1/8... Discriminator Loss: 0.6466... Generator Loss: 5.4141
Epoch 1/8... Discriminator Loss: 0.4500... Generator Loss: 4.1136
Epoch 1/8... Discriminator Loss: 0.4939... Generator Loss: 2.2410
Epoch 1/8... Discriminator Loss: 0.4472... Generator Loss: 4.8762
Epoch 1/8... Discriminator Loss: 0.5895... Generator Loss: 1.9577
Epoch 1/8... Discriminator Loss: 0.5404... Generator Loss: 3.7620
Epoch 1/8... Discriminator Loss: 0.4542... Generator Loss: 3.4601
Epoch 1/8... Discriminator Loss: 1.0423... Generator Loss: 5.1043
Epoch 1/8... Discriminator Loss: 0.5499... Generator Loss: 4.8359
Epoch 1/8... Discriminator Loss: 0.4407... Generator Loss: 2.7755
Epoch 1/8... Discriminator Loss: 1.0862... Generator Loss: 3.6536
Epoch 1/8... Discriminator Loss: 0.4142... Generator Loss: 3.6067

Epoch 1/8... Discriminator Loss: 0.4129... Generator Loss: 3.0315
Epoch 1/8... Discriminator Loss: 0.4338... Generator Loss: 4.6359
Epoch 1/8... Discriminator Loss: 0.4952... Generator Loss: 4.5443
Epoch 1/8... Discriminator Loss: 0.3993... Generator Loss: 4.6362
Epoch 1/8... Discriminator Loss: 0.4902... Generator Loss: 2.2923
Epoch 1/8... Discriminator Loss: 1.9589... Generator Loss: 2.1767
Epoch 1/8... Discriminator Loss: 1.0881... Generator Loss: 0.7289
Epoch 1/8... Discriminator Loss: 0.9587... Generator Loss: 1.7928
Epoch 1/8... Discriminator Loss: 0.5443... Generator Loss: 1.9360
Epoch 1/8... Discriminator Loss: 0.8142... Generator Loss: 3.2917
Epoch 1/8... Discriminator Loss: 0.5623... Generator Loss: 1.8491
Epoch 1/8... Discriminator Loss: 0.4716... Generator Loss: 2.7698
Epoch 1/8... Discriminator Loss: 0.5763... Generator Loss: 3.9523
Epoch 1/8... Discriminator Loss: 0.5294... Generator Loss: 3.5996
Epoch 1/8... Discriminator Loss: 1.3027... Generator Loss: 0.5989
Epoch 1/8... Discriminator Loss: 0.8998... Generator Loss: 2.2830
Epoch 1/8... Discriminator Loss: 0.8102... Generator Loss: 1.2048
Epoch 1/8... Discriminator Loss: 0.5956... Generator Loss: 3.5082
Epoch 1/8... Discriminator Loss: 0.5778... Generator Loss: 3.1179
Epoch 1/8... Discriminator Loss: 0.5883... Generator Loss: 1.7428

Epoch 1/8... Discriminator Loss: 0.9300... Generator Loss: 3.7541
Epoch 1/8... Discriminator Loss: 0.4426... Generator Loss: 2.7987
Epoch 1/8... Discriminator Loss: 0.4967... Generator Loss: 3.8464
Epoch 1/8... Discriminator Loss: 0.5121... Generator Loss: 2.2536
Epoch 1/8... Discriminator Loss: 0.4876... Generator Loss: 3.2195
Epoch 1/8... Discriminator Loss: 1.1898... Generator Loss: 4.1403
Epoch 1/8... Discriminator Loss: 0.4119... Generator Loss: 3.3613
Epoch 1/8... Discriminator Loss: 0.9849... Generator Loss: 0.8731
Epoch 1/8... Discriminator Loss: 0.4726... Generator Loss: 3.7029
Epoch 1/8... Discriminator Loss: 0.5996... Generator Loss: 1.7259
Epoch 1/8... Discriminator Loss: 0.6439... Generator Loss: 4.5052
Epoch 1/8... Discriminator Loss: 0.4560... Generator Loss: 3.7598
Epoch 1/8... Discriminator Loss: 0.4678... Generator Loss: 4.0682
Epoch 1/8... Discriminator Loss: 0.4653... Generator Loss: 2.6815
Epoch 1/8... Discriminator Loss: 0.4827... Generator Loss: 4.2681
Epoch 1/8... Discriminator Loss: 0.4117... Generator Loss: 2.9292
Epoch 1/8... Discriminator Loss: 0.4819... Generator Loss: 4.3004
Epoch 1/8... Discriminator Loss: 0.4230... Generator Loss: 2.8933
Epoch 1/8... Discriminator Loss: 1.2804... Generator Loss: 4.8427
Epoch 1/8... Discriminator Loss: 0.9281... Generator Loss: 1.0143

Epoch 1/8... Discriminator Loss: 0.5550... Generator Loss: 3.6460
Epoch 1/8... Discriminator Loss: 0.5078... Generator Loss: 4.1913
Epoch 1/8... Discriminator Loss: 0.5156... Generator Loss: 2.2364
Epoch 1/8... Discriminator Loss: 0.6573... Generator Loss: 4.5262
Epoch 1/8... Discriminator Loss: 0.5980... Generator Loss: 1.7117
Epoch 1/8... Discriminator Loss: 0.9525... Generator Loss: 4.3067
Epoch 1/8... Discriminator Loss: 0.4722... Generator Loss: 2.3767
Epoch 1/8... Discriminator Loss: 0.5176... Generator Loss: 2.9059
Epoch 1/8... Discriminator Loss: 0.9842... Generator Loss: 4.7592
Epoch 1/8... Discriminator Loss: 0.9397... Generator Loss: 3.1482
Epoch 1/8... Discriminator Loss: 0.4914... Generator Loss: 2.6215
Epoch 1/8... Discriminator Loss: 0.6049... Generator Loss: 4.4341
Epoch 1/8... Discriminator Loss: 0.4691... Generator Loss: 2.4438
Epoch 1/8... Discriminator Loss: 0.8883... Generator Loss: 4.3220
Epoch 1/8... Discriminator Loss: 0.9928... Generator Loss: 0.9096
Epoch 1/8... Discriminator Loss: 0.6563... Generator Loss: 3.2429
Epoch 1/8... Discriminator Loss: 0.5180... Generator Loss: 3.6563
Epoch 1/8... Discriminator Loss: 0.6321... Generator Loss: 2.7134
Epoch 1/8... Discriminator Loss: 0.4407... Generator Loss: 2.9092
Epoch 1/8... Discriminator Loss: 1.0334... Generator Loss: 3.7730

Epoch 1/8... Discriminator Loss: 0.7146... Generator Loss: 1.3797
Epoch 1/8... Discriminator Loss: 0.5454... Generator Loss: 3.0765
Epoch 1/8... Discriminator Loss: 1.2033... Generator Loss: 3.2457
Epoch 1/8... Discriminator Loss: 0.6888... Generator Loss: 3.0996
Epoch 1/8... Discriminator Loss: 0.5705... Generator Loss: 1.8259
Epoch 1/8... Discriminator Loss: 0.6655... Generator Loss: 2.9355
Epoch 1/8... Discriminator Loss: 0.5449... Generator Loss: 2.4215
Epoch 1/8... Discriminator Loss: 0.5159... Generator Loss: 2.2365
Epoch 1/8... Discriminator Loss: 0.6564... Generator Loss: 2.9884
Epoch 1/8... Discriminator Loss: 0.6210... Generator Loss: 1.5751
Epoch 1/8... Discriminator Loss: 0.9532... Generator Loss: 0.9012
Epoch 1/8... Discriminator Loss: 0.5823... Generator Loss: 1.7672
Epoch 1/8... Discriminator Loss: 1.0438... Generator Loss: 2.9927
Epoch 1/8... Discriminator Loss: 2.0081... Generator Loss: 0.2494
Epoch 1/8... Discriminator Loss: 0.8761... Generator Loss: 1.9917
Epoch 1/8... Discriminator Loss: 0.7482... Generator Loss: 1.2779
Epoch 1/8... Discriminator Loss: 0.6089... Generator Loss: 2.9793
Epoch 1/8... Discriminator Loss: 0.5026... Generator Loss: 2.2701
Epoch 1/8... Discriminator Loss: 0.5912... Generator Loss: 2.9213
Epoch 1/8... Discriminator Loss: 0.6389... Generator Loss: 3.9659

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.